Few-shot Text Classification with Dual Contrastive Consistency
This work improves few-shot text classification for NLP applications, but it is incremental as it builds on existing contrastive learning and consistency regularization techniques.
The paper tackles few-shot text classification by addressing overfitting from cross-entropy loss, using supervised contrastive learning on labeled data and consistency regularization on unlabeled data, resulting in a model (FTCC) that outperforms state-of-the-art methods with better robustness across four datasets.
In this paper, we explore how to utilize pre-trained language model to perform few-shot text classification where only a few annotated examples are given for each class. Since using traditional cross-entropy loss to fine-tune language model under this scenario causes serious overfitting and leads to sub-optimal generalization of model, we adopt supervised contrastive learning on few labeled data and consistency-regularization on vast unlabeled data. Moreover, we propose a novel contrastive consistency to further boost model performance and refine sentence representation. After conducting extensive experiments on four datasets, we demonstrate that our model (FTCC) can outperform state-of-the-art methods and has better robustness.